yep, you're right. I fixed the decoder to output this for confusion network
input:
...
PhraseModel I 0.50
PhraseModel tm 0.20
PhraseModel tm 0.20
PhraseModel tm 0.20
PhraseModel tm 0.20
PhraseModel tm 0.20
https://github.com/moses-smt/mosesdecoder/commit/44a0e
This is a basic question as I am relatively new to Moses.
Can someone tell me why the alignment of texts is not picking up many (common)
words and phrases from the input? The decoder shows many UNKs that should not
be.
I am experimenting with a factored model using EMS and parallel corpora from
> On 02/04/2013 20:42, Rob Voigt wrote:
> > Hi all,
> >
> > I'm trying to use confusion networks with Moses - I can decode with them
successfully, manually inputting
> an insertion weight (-weight-i), but am running into trouble doing MERT on
a model that uses confusion networks.
> >
> > Running
Hi,
"dot" is a program that is used to create the workflow image.
It usually comes with the graphviz library, so you would have
to install that.
-phi
On Sat, Apr 20, 2013 at 11:44 AM, Saeed Farzi wrote:
> Dear all,
>
> When i try to run moses example for Hierarchical phrase-based with
> EMS, I
Dear all,
When i try to run moses example for Hierarchical phrase-based with
EMS, I faced with below errors. any body can help me ? tnx
saeedfarzi@ubuntuSAEED:~/nlp/ex$
/home/saeedfarzi/nlp/moses2012/scripts/ems/experiment.perl -config
/home/saeedfarzi/nlp/moses2012/scripts/ems/example/config.hi